HeatMap V2: Wallet Behavior Analytics Platform

An interactive blockchain wallet behavior analysis platform powered by machine learning, designed for Web3 marketers to identify wallet patterns, communities, and connections.

Features

Advanced ML Analysis

• t-SNE & PCA Visualization: Dimensionality reduction for exploring wallet behavior clusters
• Behavior-Based Communities: Automatic classification into 9 distinct communities:
  • NFT Trader (127 wallets)
  • Stablecoin User (65 wallets)
  • Meme Trader (2 wallets)
  • DeFi User (1 wallet)
  • Receiver/Collector (195 wallets)
  • Sender/Distributor (58 wallets)
  • High Activity Hub (14 wallets)
  • Mixed Behavior (46 wallets)
  • Wrapper/Bridge User (3 wallets)

Interactive Visualizations

• Scatter Plot: PCA/t-SNE embeddings with community color coding and clickable wallets
• Network Graph: Wallet-to-wallet interaction network with force-directed layout
• Community Statistics: Detailed breakdown of each behavior community
• Interactive Filters: Filter by community, toggle visualization methods
• Wallet Search: Search for specific wallet addresses
• Export Functionality: Export filtered wallet lists in CSV, JSON, or TXT formats

Data Coverage

• 511 analyzed wallets from minimal dataset
• 21,181 total transfers (ERC20, ERC721, ERC1155)
• 10,214 network nodes with 12,363 connections
• 9 behavior communities detected

Quick Start

Prerequisites

• Python 3.9+
• Node.js 18+
• pip and npm

Backend Setup

cd backend

# Install dependencies
pip install -r requirements.txt

# Run ML analysis on minimal data
python visualize.py

# Start API server
python -m uvicorn api:app --host 0.0.0.0 --port 8000 --reload

Backend will be available at http://localhost:8000

Frontend Setup

cd frontend

# Install dependencies
npm install

# Start development server
npm run dev

Frontend will be available at http://localhost:3000

API Endpoints

Visualization Endpoints

• GET /api/visualization - Complete visualization data (wallets, embeddings, network, stats)
• GET /api/communities - Behavior community statistics
• GET /api/embeddings?method=tsne&community=NFT+Trader - Filtered embeddings
• GET /api/wallets - Paginated wallet list
• GET /api/stats - Overall statistics
• GET /health - Health check

Behavior Community Detection

Classification Logic

• NFT Trader: Wallets with ERC721/ERC1155 transactions
• Stablecoin User: >50% stablecoin (USDT, USDC, DAI) transactions
• Meme Trader: >50% meme coin (PEPE, SHIB, etc.) transactions
• DeFi User: >50% DeFi protocol (UNI, AAVE, etc.) transactions
• Wrapper/Bridge User: >50% wrapped asset (WETH, WBTC) transactions
• High Activity Hub: >20 unique counterparties
• Sender/Distributor: >80% outgoing transactions
• Receiver/Collector: >80% incoming transactions
• Mixed Behavior: No dominant pattern

Use Cases for Web3 Marketers

1. Audience Segmentation

Identify and target specific wallet behaviors:

• NFT collectors for NFT project launches
• Stablecoin users for trading platforms
• DeFi users for protocol integrations
• Export wallet lists filtered by community for outreach campaigns

2. Wallet Clustering

Discover similar wallet patterns using t-SNE visualization:

• Nearby wallets in embedding space have similar behaviors
• Filter by community to focus on specific segments
• Click on any wallet to view detailed behavior breakdown

3. Wallet Research

Search and analyze specific wallets:

• Search by wallet address to find specific targets
• View comprehensive wallet details (transactions, token preferences, behavior ratios)
• Copy addresses for use in other tools or campaigns
• Export to Etherscan or DeBank for deeper analysis

4. Network Analysis

Find connected wallets:

• Identify wallet clusters that transact together
• Discover influential hubs with high connectivity
• Map token distribution networks

5. Data Export for Campaigns

Export wallet lists for marketing use:

• CSV export: Full wallet data with metrics and community labels
• JSON export: Structured data for programmatic use
• TXT export: Address-only lists for easy copy-paste into marketing tools

Technology Stack

Backend

• FastAPI, Pandas, Scikit-learn (t-SNE, PCA, KMeans)
• NetworkX for graph analysis
• Web3.py, Etherscan API, The Graph

Frontend

• Next.js 14 + TypeScript
• Recharts for scatter plots
• Canvas API for network graphs
• Tailwind CSS

License

MIT